Harshitha Machiraju

I am a Machine Learning Researcher focused on enhancing the robustness of neural networks by evaluating and addressing challenges posed by adversarial perturbations, image corruptions, and background variations. Recently awarded a PhD from EPFL under the supervision of Prof. Pascal Frossard and Prof. Michael Herzog, I have developed various methodologies to uncover biases and fortify the robustness of deep neural networks to distribution shifts.